36 replies

1. mortendk 🤘 3 min later

   @zachleat dont you ever sleep ;)

   or is it the post-musk-twitter-rush thats kicking in ?

   1. Zach Leatherman 13 min later

      @mortendk I definitely have peaks and slumps of momentum—I try to ride them on the upswing when possible!

      1. mortendk 🤘 19 min later

         @zachleat

         [insert gif with surf dude johnny utah aka keanu Reeves saying inspirational quote about surfing the waves of inspiration

         1. Zach Leatherman 26 min later

            @mortendk cowabunga

2. Torsten Knabe 4 min later

   @zachleat If only I could request my archives 😭 It's an endless loop where I need to put in my password and then get a verification token from an email but my guess is a shut off microservice sends that email so I never get the code

   1. Zach Leatherman 12 min later

      @torkna oh no! I don’t think it’s 100% failing, I’ve made it through the email verification token workflow at least once this week!

      1. Torsten Knabe 24 min later

         @zachleat ooh good to know, I will try again and see what happens. Thanks!

         1. Zach Leatherman 25 min later

            @torkna (also I’m sure you’ve done this but just in case—check your spam folder!)

3. dominik schwind 13 min later

   @zachleat it might be a good idea to separate the implementation from the database. This way other implementations in other languages can use the same canonical list.

   1. Zach Leatherman 15 min later

      @dominik Great idea! Filed this here https://github.com/tweetback/tweetback-canonical/issues/2

4. dan leatherman 23 min later

   @zachleat seoh my gosh how do you even think of these things

   1. Zach Leatherman 26 min later

      @danleatherman brain do stuff, sometimes

5. Stuart Langridge 34 min later

   @zachleat ooh, I was thinking about building something to turn my Twitter export into a useful website, and now I see you're already on the case! Nice. I shall use that :)

   1. Zach Leatherman 36 min later

      @sil yeah—working with @nhoizey on this!

      you can see a few previews I’ve shipped here:
      https://www.zachleat.com/twitter/
      https://twitter.11ty.dev

      1. Andy Bell 42 min later

         @zachleat @sil @nhoizey this looks great!

      2. ge ricci 45 min later

         @zachleat @sil @nhoizey this is awesome!

6. François REMY 1 hr later

   @zachleat Cool! Did you redownload all your tweets yourself? Because the archives has many issues... Did you pay for an API token?

   1. Zach Leatherman 1 hr later

      @FremyCompany I used the archive as a baseline and then repopulate from the API—I wasn’t aware API tokens costed money 👀

      1. François REMY 2 hr later

         @zachleat Oh, I don't know if they cost anything, I'm just asking for your experience here to learn more about what you did.

         Do you care for instance about your retweets, or just your own tweets?

         I find that backing up your retweets / quoted tweets / likes is the most difficult, I was wondering if you had any thoughts on this already.

         1. Zach Leatherman 2 hr later

            @FremyCompany I do back up retweets, although the content is truncated e.g. https://www.zachleat.com/twitter/1587076514939641856/

            quote tweets link directly to the source https://www.zachleat.com/twitter/1586436147386888193/

            My concern with the retweet thing is that they don’t really obey norms around upstream deletes

            1. François REMY 2 hr later

               @zachleat Yes, because of the truncation (which often deletes the attached media and expanded urls), I am currently recrawling all of them myself (masquerading as the twitter website on my PC).

               1. François REMY 2 hr later

                  @zachleat But for now, I'm only thinking about keeping that data private, but it's interesting to think about the issue of keeping backups of other people's messages that they can't delete easily.

                  I will keep this in mind if I want to host this later on, thanks for the thought!

                  Any other challenge like this you encountered?

               2. Zach Leatherman 2 hr later

                  @FremyCompany maybe we should collab on tweetback! We have about 8 folks on the org right now

                  cc @nhoizey

                  1. Nicolas Hoizey 2 hr later

                     @zachleat I think we should open the repository ASAP, and issues and PRs will be a good way to discuss such topics.
                     @fremycompany

                     1. Zach Leatherman 2 hr later

                        @nhoizey @FremyCompany huge upvote—I think it’s ready enough for early previews

                        1. Nicolas Hoizey 2 hr later

                           @zachleat I think so!

                           @fremycompany

7. Zach Leatherman 1 hr later

   You can see this in action on the Eleventy Twitter Archive, where any links to https://twitter.com/zachleat/ now link to https://zachleat.com/twitter/ instead.

   e.g. https://twitter.11ty.dev/1238956883488235520/

   1. Bryan Schuetz 1 hr later

      @zachleat This is awesome. Now can you build something that will preserve all the t.co shortened links that will disappear when the whole thing goes whale-up?

      1. Zach Leatherman 2 hr later

         @zero_gravitas tweetback archives will get rid of t.co links in your content when they build! 🏆

         1. Bryan Schuetz 2 hr later

            @zachleat 🤩

   2. Adrianna Tan 2 hr later

      @zachleat thanks for doing this. Zach. I was planning to learn 11ty (mostly Hugo at the moment) to put together my archives so this is most helpful

      1. Zach Leatherman 2 hr later

         @skinnylatte yeah! Stay tuned! We will have a starter project here for this in a bit!

         https://fosstodon.org/@eleventy/109379173289876463

8. tea 🌺 5 hr later

   @zachleat You may also want to consider transforming all twitter links in your archive to nitter https://nitter.snopyta.org/zachleat, for example rather than linking to https://twitter.com/buildWithLit/status/1591173922678079489 you can link to https://nitter.snopyta.org/buildWithLit/status/1591173922678079489 details about nitter https://nitter.snopyta.org/about, this can act as a fallback if they haven't created their own archive

   1. Zach Leatherman 5 hr later

      @thomasreggi ooh can you file an issue on the repo for this? I think it’s worth discussing

      1. tea 🌺 5 hr later

         @zachleat done https://github.com/tweetback/tweetback-canonical/issues/3

         1. Zach Leatherman 5 hr later

            @thomasreggi Thank you!!